
# Data handling
import pandas as pd
import numpy as np
import datetime
from math import sqrt
from numpy import loadtxt
from itertools import product
from tqdm import tqdm

# Visualisations
!pip install chart_studio
import seaborn as sns
import matplotlib.pyplot as plt
import chart_studio.plotly as py
import plotly.graph_objs as go
from plotly.offline import init_notebook_mode, iplot
from scipy import stats
from wordcloud import WordCloud
from wordcloud import STOPWORDS

# Machine learning
import xgboost as xgb
from sklearn import preprocessing
from xgboost import plot_tree
from sklearn.metrics import accuracy_score
from sklearn.metrics import mean_squared_error
from sklearn.preprocessing import LabelEncoder
from sklearn.model_selection import KFold
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.model_selection import train_test_split
from sklearn.metrics import precision_score
from sklearn.metrics import recall_score
from lightgbm import LGBMRegressor

Requirement already satisfied: chart_studio in c:\users\krish\anaconda3\lib\site-packages (1.1.0)
Requirement already satisfied: plotly in c:\users\krish\anaconda3\lib\site-packages (from chart_studio) (5.9.0)
Requirement already satisfied: retrying>=1.3.3 in c:\users\krish\anaconda3\lib\site-packages (from chart_studio) (1.3.3)
Requirement already satisfied: requests in c:\users\krish\anaconda3\lib\site-packages (from chart_studio) (2.28.1)
Requirement already satisfied: six in c:\users\krish\anaconda3\lib\site-packages (from chart_studio) (1.16.0)
Requirement already satisfied: tenacity>=6.2.0 in c:\users\krish\anaconda3\lib\site-packages (from plotly->chart_studio) (8.0.1)
Requirement already satisfied: charset-normalizer<3,>=2 in c:\users\krish\anaconda3\lib\site-packages (from requests->chart_studio) (2.0.4)
Requirement already satisfied: idna<4,>=2.5 in c:\users\krish\anaconda3\lib\site-packages (from requests->chart_studio) (3.3)
Requirement already satisfied: certifi>=2017.4.17 in c:\users\krish\anaconda3\lib\site-packages (from requests->chart_studio) (2022.9.24)
Requirement already satisfied: urllib3<1.27,>=1.21.1 in c:\users\krish\anaconda3\lib\site-packages (from requests->chart_studio) (1.26.11)


training_data = pd.read_csv("sales_train.csv",parse_dates = ['date'], infer_datetime_format = True, dayfirst = True)
test = pd.read_csv("test.csv")
items_data = pd.read_csv("items.csv")
shops_data = pd.read_csv("shops.csv")
item_categories = pd.read_csv("item_categories.csv")
sample_submission = pd.read_csv("sample_submission.csv")


# Showcase the initial 5 rows of the training_data dataset which needs to be modelled for future prediction
training_data.head()


# Showcase the initial 5 rows of the test dataset
# It contains item_id, shop_id
# Need to predict item_cnt_day
test.head()


# Showcase the initial 5 rows of the items dataset
# It contains item_name, item_id, item_category_id
items_data.head()


# Showcase the initial 5 rows of the shops dataset
# It contains shop_name, shop_id
shops_data.head()


# Showcase the initial 5 rows of the items_categories dataset
# It contains item_category_name and id
item_categories.head()


# Showcase the initial 5 rows of sample submission dataset
sample_submission.head()


#Size of all the datasets

print("Size of Training Dataset: ", training_data.shape)
print("Size of Testing Dataset: ", test.shape)
print("Size of Items Dataset: ", items_data.shape)
print("Size of Shops Dataset: ", shops_data.shape)
print("Size of Item Categories Dataset, ", item_categories.shape)
print("Size of Sample Submssion Dataset: ", sample_submission.shape)

Size of Training Dataset:  (2935849, 6)
Size of Testing Dataset:  (214200, 3)
Size of Items Dataset:  (22170, 3)
Size of Shops Dataset:  (60, 2)
Size of Item Categories Dataset,  (84, 2)
Size of Sample Submssion Dataset:  (214200, 2)


# Merging all the files with common parameters as shop_id, item_id, item_category_id
df = pd.merge(training_data, shops_data, how = 'left', left_on = 'shop_id', right_on = 'shop_id')
df = pd.merge(df, items_data, how = 'left', left_on = 'item_id', right_on = 'item_id')
df = pd.merge(df, item_categories, how = 'left', left_on = 'item_category_id', right_on = 'item_category_id')
df.drop(['item_id', 'item_category_id', 'shop_id'], axis = 1, inplace = True)
df.head()


# The information of the dataframe created by merging the datasets
# The datatypes of the attributes are shown specifically
df.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 2935849 entries, 0 to 2935848
Data columns (total 7 columns):
 #   Column              Dtype         
---  ------              -----         
 0   date                datetime64[ns]
 1   date_block_num      int64         
 2   item_price          float64       
 3   item_cnt_day        float64       
 4   shop_name           object        
 5   item_name           object        
 6   item_category_name  object        
dtypes: datetime64[ns](1), float64(2), int64(1), object(3)
memory usage: 179.2+ MB


# Size of merged dataset before removing duplicates
df.shape

(2935849, 7)


# Check for duplicates
df.duplicated().sum()

6


# Remove the duplicates
df.drop_duplicates(keep = 'first', inplace = True)


# Size of merged dataset after removing duplicates
df.shape

(2935843, 7)


# Basic statistical information about the data (mean, sd, min, max)
df.describe()


# Explore the distbution sales data
data = training_data[['shop_id','item_price', 'item_cnt_day']]
def num_plot(df, var):
    fig = plt.figure(figsize = (16, 5))
    plt.subplot(1,3,1)
    df[var].hist(bins = 40)
    plt.title(f"Distribution of {var}")
    
    plt.subplot(1,3,2)
    stats.probplot(df[var], dist = "norm", plot = plt)
    plt.ylabel('Quantiles')
    
    plt.subplot(1,3,3)
    sns.boxplot(y = df[var],color="red")
    plt.title("Boxplot")
    plt.show()
    
for var in data:
    num_plot(data, var)


# Make a plot to show the frequence of sales among shops
plt.figure(figsize = (20, 20))
sns.countplot(y=df['shop_name'], palette = ("Greens"))
plt.xlabel('Number of sold items in each shop')
plt.ylabel('Name of the shop')

Text(0, 0.5, 'Name of the shop')


# Show shop daily sales
shop_daily_sales = training_data.groupby(['shop_id', 'date_block_num'], as_index = False)['item_cnt_day'].sum()

shop_daily_sales_sc = []
for shop in shop_daily_sales['shop_id'].unique():
    current_shop_daily_sales = shop_daily_sales[(shop_daily_sales['shop_id']) == shop]
    shop_daily_sales_sc.append(go.Scatter(x=current_shop_daily_sales['date_block_num'], y = current_shop_daily_sales['item_cnt_day'], name=('shop_id %s' % shop)))
    
layout = go.Layout(title='Shop daily sales', xaxis=dict(title='Time (2013 - 2015)'), yaxis=dict(title='Sales'))
fig = go.Figure(data = shop_daily_sales_sc, layout=layout)
iplot(fig)


# Output the first 10 best selling items
best_selling_items = df['item_name'].value_counts().head(10)
best_selling_items = best_selling_items.to_frame().reset_index()
best_selling_items.columns = ['item_name', 'item_count']
best_selling_items.head(10)


# Plotting a graph
fig = plt.figure(figsize = (20, 10))
sns.barplot(x=best_selling_items['item_count'], y=best_selling_items['item_name'], data=best_selling_items)
plt.xlabel('Item Count')
plt.ylabel('Item Name')

Text(0, 0.5, 'Item Name')


# Output the first 10 best selling categories
best_selling_categories = df['item_category_name'].value_counts().head(10)
best_selling_categories = best_selling_categories.to_frame().reset_index()
best_selling_categories.columns = ['item_category_name', 'count']
best_selling_categories.head(10)


# Plotting a graph
fig = plt.figure(figsize = (20, 10))
sns.barplot(x = best_selling_categories['count'], y = best_selling_categories['item_category_name'], data = best_selling_categories)
plt.xlabel('Category Count')
plt.ylabel('Item Name')

Text(0, 0.5, 'Item Name')


# Making a word cloud for shop name

plt.rcParams['figure.figsize'] = (5, 5)
stopwords = set(STOPWORDS)
wordcloud1 = WordCloud(background_color = 'pink',
                      max_words = 200, 
                      stopwords = stopwords,
                     width = 1000,
                     height = 1000,
                     random_state = 42).generate(str(shops_data['shop_name']))


plt.title('Wordcloud for Shop Names', fontsize = 30)
plt.axis('off')
plt.imshow(wordcloud1, interpolation = 'bilinear')

<matplotlib.image.AxesImage at 0x23992160760>


# Making a word cloud for item name

plt.rcParams['figure.figsize'] = (5, 5)
stopwords = set(STOPWORDS)
wordcloud2 = WordCloud(background_color = 'lightgreen',
                      max_words = 200, 
                      stopwords = stopwords,
                     width = 1000,
                     height = 1000,
                     random_state = 42).generate(str(items_data['item_name']))


plt.title('Wordcloud for Item Names', fontsize = 30)
plt.axis('off')
plt.imshow(wordcloud2, interpolation = 'bilinear')

<matplotlib.image.AxesImage at 0x239a35eb9a0>


# Making a word cloud for item categories name

plt.rcParams['figure.figsize'] = (5, 5)
stopwords = set(STOPWORDS)
wordcloud3 = WordCloud(background_color = 'lightblue',
                      max_words = 200, 
                      stopwords = stopwords,
                     width = 1000,
                     height = 1000,
                     random_state = 42).generate(str(item_categories['item_category_name']))


plt.title('Wordcloud for Item Category Names', fontsize = 30)
plt.axis('off')
plt.imshow(wordcloud3, interpolation = 'bilinear')

<matplotlib.image.AxesImage at 0x239924c02b0>


# Calculate IQR
Q1 = np.percentile(df.item_price,25)/100
Q3 = np.percentile(df.item_price,75)/100
IQR = Q3-Q1

# Drop item_price 1.5 times IQR above the below the max and min item_price
# Drop item_cnt_day with standard deviation higher than 3 
training_data.drop(training_data[training_data.item_price > 1.5*IQR + training_data.item_price.max()].index, inplace = True)
training_data.drop(training_data[training_data.item_price < 1.5*IQR + training_data.item_price.min()].index, inplace = True)
training_data.drop(training_data[training_data.item_cnt_day > 3 * training_data.item_cnt_day.std()].index, inplace = True)

# There are a few negative item_counts
# Thus we need to eliminate negative values
training_data.drop(training_data[training_data.item_price < 0].index, inplace = True)
training_data.drop(training_data[training_data.item_cnt_day < 0].index, inplace = True)


# Calculate revenue
training_data['revenue'] = training_data.item_price * training_data.item_cnt_day


# Group data based on shop id and item id then get value of sales per month
pivot_training_data = training_data.pivot_table(index=['shop_id', 'item_id'], columns='date_block_num', values='item_cnt_day', aggfunc='sum').fillna(0.0)
pivot_training_data.head(20)


# Size of data after removing outliers
print("Size of Training Dataset after removing outliers: ",training_data.shape)

# Size of data after creating pivot table
print("Size of Pivot Dataset: ",pivot_training_data.shape)

Size of Training Dataset after removing outliers:  (2874701, 7)
Size of Pivot Dataset:  (423533, 34)


# Data Cleaning
training_df = pivot_training_data.reset_index()
training_df["shop_id"] = training_df.shop_id.astype("str")
training_df["item_id"] = training_df.item_id.astype("str")


# Merge the items with categories
item_cat_df = items_data.merge(item_categories[['item_category_id','item_category_name']], how="inner", on="item_category_id")[['item_id','item_category_name']]
item_cat_df['item_id'] = item_cat_df.item_id.astype('str')

# Merge the train dataframe with the items and categories dataframe
training_df = training_df.merge(item_cat_df, how="inner", on="item_id")


# Perform Label encoding
number = preprocessing.LabelEncoder()
# Check the variables in the train dataframe
training_df


# Encode the column with categorical data
training_df['item_category_name'] = number.fit_transform(training_df['item_category_name'])


# Rearrange the columns
training_df = training_df[['shop_id', 'item_id', 'item_category_name'] + list(range(34))]
training_df.head()


# Modelling
param = {'max_depth': 13, 
         'subsample':1,
         'min_child_weight': 0.7,
         'eta':0.3, 
         'num_round':1000, 
         'seed':1,
         'silent':0,
         'eval_metric':'rmse',
         'n_estimators': 120}

# Generate a DMatrix to train the model
xgbtrain = xgb.DMatrix(training_df.iloc[:,  (training_df.columns != 33)].values, training_df.iloc[:, training_df.columns == 33].values)
watchlist  = [(xgbtrain,'train-rmse')]

# Train the model
bst = xgb.train(param, xgbtrain)
preds = bst.predict(xgb.DMatrix(training_df.iloc[:,  (training_df.columns != 33)].values))
rmse = np.sqrt(mean_squared_error(preds,training_df.iloc[:, training_df.columns == 33].values))

# Print the RMSE of the model
print("RMSE Value: ",rmse)

[16:56:06] WARNING: C:/buildkite-agent/builds/buildkite-windows-cpu-autoscaling-group-i-03de431ba26204c4d-1/xgboost/xgboost-ci-windows/src/learner.cc:767: 
Parameters: { "n_estimators", "num_round", "silent" } are not used.

RMSE Value:  0.4302462208714635


# See the the most important features
fig, ax = plt.subplots(figsize=(20, 10))
xgb.plot_importance(bst, ax = ax, importance_type='gain')

<AxesSubplot:title={'center':'Feature importance'}, xlabel='F score', ylabel='Features'>


# Preparing the test data frame
test_df = test
test_df.shop_id = test_df.shop_id.astype('str')
test_df.item_id = test_df.item_id.astype('str')


test_df = test.merge(training_df, how="left", on=["shop_id", "item_id"]).fillna(0.0)
# Check the first five variables of the test dataframe
test_df.head()


d = dict(zip(test_df.columns[4:],list(np.array(list(test_df.columns[4:])) - 1)))

test_df = test_df.rename(d, axis = 1)


# Make sales predictions for the test data
predictions = bst.predict(xgb.DMatrix(test_df.iloc[:, (test_df.columns != 'ID') & (test_df.columns != -1)].values))


predictions = list(map(lambda x: min(20, max(x,0)), list(predictions)))
submission = pd.DataFrame({'ID': test_df.ID, 'item_cnt_month':predictions})
submission['item_cnt_month']  = round(submission['item_cnt_month'] , 2)


submission.to_csv('submission1.csv', index=False)


print(submission)

            ID  item_cnt_month
0            0            0.46
1            1            0.15
2            2            0.32
3            3            0.13
4            4            0.15
...        ...             ...
214195  214195            0.14
214196  214196            0.13
214197  214197            0.05
214198  214198            0.13
214199  214199            0.05

[214200 rows x 2 columns]


# Make a dataset with only monthly sales data
data = training_data.groupby([training_data['date'].apply(lambda x: x.strftime('%Y-%m')),'item_id','shop_id']).sum().reset_index()

# Specify the important attributes which we want to add to the data
data = data[['date','item_id','shop_id','item_cnt_day']]

# Group data based on shop id and item id then get value of sales per month
data = data.pivot_table(index=['item_id','shop_id'], columns = 'date', values = 'item_cnt_day', fill_value = 0).reset_index()

data.shape

(423533, 36)


# Merge the monthly sales data prepared to the test data set
test = pd.read_csv("test.csv")
test = pd.merge(test, data, on = ['item_id', 'shop_id'], how = 'left')

# Filling the empty values found in the dataset
test.fillna(0, inplace = True)
test.head()


# Create the actual training data

x_train = test.drop(['2015-10', 'item_id', 'shop_id'], axis = 1)
y_train = test['2015-10']

# Deleting the first column so that it can predict the future sales data
x_test = test.drop(['2013-01', 'item_id', 'shop_id'], axis = 1)

print("Shape of x_train :", x_train.shape)
print("Shape of x_test :", x_test.shape)
print("Shape of y_train :", y_train.shape)

Shape of x_train : (214200, 34)
Shape of x_test : (214200, 34)
Shape of y_train : (214200,)


# Splitting the data into train and valid dataset

x_train, x_valid, y_train, y_valid = train_test_split(x_train, y_train, test_size = 0.2, random_state = 0)

print("Shape of x_train :", x_train.shape)
print("Shape of x_valid :", x_valid.shape)
print("Shape of y_train :", y_train.shape)
print("Shape of y_valid :", y_valid.shape)

Shape of x_train : (171360, 34)
Shape of x_valid : (42840, 34)
Shape of y_train : (171360,)
Shape of y_valid : (42840,)


# MODELING
model_lgb = LGBMRegressor( n_estimators=200,
                           learning_rate=0.03,
                           num_leaves=32,
                           colsample_bytree=0.9497036,
                           subsample=0.8715623,
                           max_depth=8,
                           reg_alpha=0.04,
                           reg_lambda=0.073,
                           min_split_gain=0.0222415,
                           min_child_weight=40)
model_lgb.fit(x_train, y_train)

y_pred_lgb = model_lgb.predict(x_test)


# Generate the model
y_pred_lgb = model_lgb.predict(x_test).clip(0., 20.)

# Create the submission file and submit
preds = pd.DataFrame(y_pred_lgb, columns=['item_cnt_month'])
preds.to_csv('submission2.csv',index_label='ID')


submission.to_csv('submission2.csv', index=False)


print(submission)

            ID  item_cnt_month
0            0            0.46
1            1            0.15
2            2            0.32
3            3            0.13
4            4            0.15
...        ...             ...
214195  214195            0.14
214196  214196            0.13
214197  214197            0.05
214198  214198            0.13
214199  214199            0.05

[214200 rows x 2 columns]

	date	shop_id	item_id	item_price	item_cnt_day
0	2013-01-02	59	22154	999.00	1.0
1	2013-01-03	25	2552	899.00	1.0
2	2013-01-05	25	2552	899.00	-1.0
3	2013-01-06	25	2554	1709.05	1.0
4	2013-01-15	25	2555	1099.00	1.0

	ID	shop_id	item_id
0	0	5	5037
1	1	5	5320
2	2	5	5233
3	3	5	5232
4	4	5	5268

	shop_name	shop_id
0	!Якутск Орджоникидзе, 56 фран	0
1	!Якутск ТЦ "Центральный" фран	1
2	Адыгея ТЦ "Мега"	2
3	Балашиха ТРК "Октябрь-Киномир"	3
4	Волжский ТЦ "Волга Молл"	4

	item_category_name	item_category_id
0	PC - Гарнитуры/Наушники	0
1	Аксессуары - PS2	1
2	Аксессуары - PS3	2
3	Аксессуары - PS4	3
4	Аксессуары - PSP	4

	date	item_price	item_cnt_day	shop_name	item_name	item_category_name
0	2013-01-02	999.00	1.0	Ярославль ТЦ "Альтаир"	ЯВЛЕНИЕ 2012 (BD)	Кино - Blu-Ray
1	2013-01-03	899.00	1.0	Москва ТРК "Атриум"	DEEP PURPLE The House Of Blue Light LP	Музыка - Винил
2	2013-01-05	899.00	-1.0	Москва ТРК "Атриум"	DEEP PURPLE The House Of Blue Light LP	Музыка - Винил
3	2013-01-06	1709.05	1.0	Москва ТРК "Атриум"	DEEP PURPLE Who Do You Think We Are LP	Музыка - Винил
4	2013-01-15	1099.00	1.0	Москва ТРК "Атриум"	DEEP PURPLE 30 Very Best Of 2CD (Фирм.)	Музыка - CD фирменного производства

DA Final Project¶

Predict Future Sales¶

Team Members: Krisha Shah, Bhavya Shah¶

Eliminating Outliers¶

Prediction Model 1¶

Prediction Model 2¶

Results:¶

Submission 1 gives a public score of 1.06217 on Kaggle competition¶

Submission 2 gives a public score of 0.87922 which leads to Rank 1629 on the leaderboard¶

	item_name	item_id	item_category_id
0	! ВО ВЛАСТИ НАВАЖДЕНИЯ (ПЛАСТ.) D	0	40
1	!ABBYY FineReader 12 Professional Edition Full...	1	76
2	***В ЛУЧАХ СЛАВЫ (UNV) D	2	40
3	***ГОЛУБАЯ ВОЛНА (Univ) D	3	40
4	***КОРОБКА (СТЕКЛО) D	4	40

	date_block_num	item_price	item_cnt_day
count	2.935843e+06	2.935843e+06	2.935843e+06
mean	1.456991e+01	8.908535e+02	1.242641e+00
std	9.422992e+00	1.729801e+03	2.618837e+00
min	0.000000e+00	-1.000000e+00	-2.200000e+01
25%	7.000000e+00	2.490000e+02	1.000000e+00
50%	1.400000e+01	3.990000e+02	1.000000e+00
75%	2.300000e+01	9.990000e+02	1.000000e+00
max	3.300000e+01	3.079800e+05	2.169000e+03

	item_name	item_count
0	Фирменный пакет майка 1С Интерес белый (34*42)...	31340
1	Playstation Store пополнение бумажника: Карта ...	9408
2	Прием денежных средств для 1С-Онлайн	9067
3	Diablo III [PC, Jewel, русская версия]	7479
4	Kaspersky Internet Security Multi-Device Russi...	6853
5	World of Warcraft. Карта оплаты игрового време...	6602
6	Grand Theft Auto V [PS3, русские субтитры]	6475
7	Call of Duty: Black Ops II [PС, Jewel, русская...	6320
8	Minecraft [Xbox 360, английская версия]	5811
9	Grand Theft Auto V [Xbox 360, русские субтитры]	5805

	item_category_name	count
0	Кино - DVD	564651
1	Игры PC - Стандартные издания	351591
2	Музыка - CD локального производства	339585
3	Игры - PS3	208219
4	Кино - Blu-Ray	192673
5	Игры - XBOX 360	146785
6	Игры PC - Дополнительные издания	121539
7	Игры - PS4	79058
8	Подарки - Мягкие игрушки	53845
9	Подарки - Настольные игры (компактные)	53227

	date_block_num	0	1	2	3	4	5	6	7	8	9	...	24	25	26	27	28	29	30	31	32	33
shop_id	item_id
0	30	0.0	22.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
	31	0.0	11.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
	32	6.0	10.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
	33	3.0	3.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
	35	1.0	14.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
	36	0.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
	40	0.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
	42	0.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
	43	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
	49	0.0	2.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
	51	2.0	3.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
	57	0.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
	59	0.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
	61	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
	75	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
	85	0.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
	88	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
	95	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
	96	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
	97	0.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0

	ID	shop_id	item_id	item_category_name	...	24	28	29	30	31	32	33
0	0	5	5037	19.0	...	2.0	1.0	1.0	1.0	3.0	1.0	0.0
1	1	5	5320	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0
2	2	5	5233	19.0	...	0.0	3.0	2.0	0.0	1.0	3.0	1.0
3	3	5	5232	23.0	...	0.0	0.0	0.0	0.0	1.0	0.0	0.0
4	4	5	5268	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0

	ID	item_cnt_month
0	0	0.5
1	1	0.5
2	2	0.5
3	3	0.5
4	4	0.5